What breaks in the months after a senior engineer leaves?
Nothing breaks on day one — that is the trap. The drawings stay, the running projects have momentum, and the team is competent. What breaks arrives in a sequence: first diagnosis speed, then design confidence, then the ability to touch legacy products at all. Each failure looks like an ordinary engineering problem; together they are the signature of a departed memory.
Weeks 1–8: the longer route to the same answer
The first symptom is speed. Questions that used to take one walk to a desk now take a meeting, a search through drawings that don't contain the answer, and an experiment. The team gets to the same conclusions — usually — but at three times the cost. Field issues on machines the veteran knew by serial number now start from zero. Nobody escalates this; it just becomes how long things take.
Months 2–8: the first repeated mistakes
Then the failure catalog starts costing money. A design choice the veteran had learned to avoid — the weld detail that cracks, the supplier whose steel disappoints, the optimization that fails in the field — gets made again by someone who never saw it fail. The first re-made mistake is usually absorbed as bad luck. The pattern only becomes visible to whoever knew the history, and he is gone.
Months 6–18: the untouchable legacy
The subtlest damage: legacy products slowly become no-go zones. Nobody remembers which margins are load-bearing and which constraints were a 2012 supplier limitation, so prudent engineers stop touching the old designs entirely — no cost reductions, no improvements, no confident quotes on modifications. The installed base, often the company's most profitable territory, freezes. The freeze looks like caution; it is actually amnesia.
What the survivors say
Ask engineering teams a year after such a departure and the phrasing is remarkably consistent: 'we keep finding out things he knew.' Not 'we miss him' — they do — but the operational version: the knowledge announces its absence one surprise at a time. The fix has the same shape as the founder case: capture before departure (see how to interview an expert), aimed at the failure catalog and the constraint history first.
Where MentX comes in
MentX is a personality-aware knowledge base: it captures the knowledge of a company's key person — from meetings and a guided intake track — into a living, temporal knowledge graph with full source attribution. MentX captures precisely the layers this sequence consumes: the diagnostic patterns and installed-base memory (Connect, from his live design reviews and service debriefs), and the failure catalog with the constraint history (Deepen, case by case). Through DigiME the team keeps asking him — 'has this been tried?', 'which margins matter here?' — with sources and dates under every answer.
